Abstract
It is not easy to ensure optimal prevention of hospital‐acquired pressure ulcer (HAPU) in crowded emergency departments (EDs). We hypothesised that a prolonged ED length of stay (LOS) is associated with an increased risk of HAPU. This is a single‐centre observational study. Prospectively collected HAPU surveillance data were analysed. Adult (aged ≥20 years) patients admitted through the ED from April 1, 2013 to December 31, 2016 were included. The primary outcome was the development of HAPU within a month. Covariates included demographics, comorbidities, conditions at triage, initial laboratory results, primary ED diagnosis, critical ED interventions, and ED dispositions. The association between ED LOS and HAPU was modelled using logistic and extended Cox regression. A total of 48 641 admissions were analysed. The crude odds ratio (OR) and hazard ratio (HR) for HAPU were increased to 1.44 (95% CI, 1.20‐1.72) and 1.21 (95% CI, 1.02‐1.45), respectively, in ED LOS ≥24 hours relative to ED LOS <6 hours. In multivariable logistic regression, ED LOS ≥12 and ≥24 hours were associated with higher risk of HAPU, with ORs of 1.30 (95% CI, 1.05‐1.60) and 1.80 (95% CI, 1.45‐2.23) relative to ED LOS <6 hours, respectively. The extended Cox regression showed that the risk lasted up to a week, with HRs of 1.42 (95% CI, 1.07‐1.88) and 1.92 (95% CI, 1.44‐2.57) relative to ED LOS <6 hours, respectively. In conclusion, Prolonged ED LOS is independently associated with HAPU. Shorter ED LOS should be pursued as a goal in a multifaceted solution for HAPU.
Keywords: electronic health records, emergency department, length of stay, pressure ulcer, surveillance
1. INTRODUCTION
A pressure ulcer is localised damage to the skin and/or underlying soft tissue usually over a bony prominence or related to a medical or other device.1 It is a significant health problem with US epidemiologic studies reporting an incidence of around 1 to 3 million per year.2
Health care‐acquired pressure ulcer (HAPU) is common, with its incidence ranging from 8.5% to 13.4% in hospitals in the United States and Canada.3 It was reported that HAPUs can lead to significant complications, including death.4 In addition, it increases the cost of care significantly, with a recent estimate of the incremental cost to hospitals of treating HAPUs of around $10 708 USD per patient and $26.8 billion USD nationally.4
With emergency department (ED) overcrowding becoming a global norm,5, 6 prolonged ED length of stay (LOS) is increasingly becoming a serious problem.7 Prolonged ED LOS has repeatedly been shown to be associated with poor outcomes in numerous conditions, including acute coronary syndrome, stroke, sepsis, and resuscitated out‐of‐hospital cardiac arrest.8
However, there is a lack of evidence of whether prolonged ED LOS is associated with increased risk of HAPU. While the ED environment may not be conducive for comprehensive preventive care to protect patients from HAPU, thus making the hypothesis very plausible, previous studies could not find any significant association between prolonged ED stay and HAPUs.9, 10, 11, 12 This might be because of the low statistical power of the studies. Considering that many ED patients are at an increased risk of HAPU, we believe that this hypothesis should be re‐evaluated with a larger dataset.13 With recent advancements in clinical informatics and hospital‐based surveillance systems, it has become possible to evaluate an association between an exposure and adverse outcomes on a larger scale. The aim of the study is to determine whether prolonged ED LOS is an independent risk factor for HAPU.
2. MATERIALS AND METHODS
2.1. Study design
This is a single‐centre observational study utilising the electronic health records (EHR) database of patients admitted through the ED from April 1, 2013 to December 31, 2016. The study facility is a tertiary academic hospital located in South Korea with an annual ED visit of over 80 000 patients a year. The institutional review boards of the facility approved the study and provided a waiver of informed consent.
2.2. Study population and the primary outcome
The study population included adult (aged ≥20) patients admitted through the ED from April 1, 2013 to December 31, 2016. Primary outcome was the occurrence of a new HAPU within a month after admission. It is possible that prolonged ED stay only accelerates the course of HAPU rather than causing new cases of HAPU, in which case any possible difference that is initially presented will diminish eventually. Therefore, we chose 1‐month incidence as our primary outcome to ensure that any observed difference is solid.
Information about new pressure ulcers (PUs) was retrieved from a PU surveillance system developed and managed by the nursing department of the hospital since April 1, 2013. The surveillance system keeps track of any new developments and progression of PUs in admitted patients and includes the time of the first recognition, anatomic location, size, stage, presumed causes, and presence of patient‐specific risk factors.
Covariates including demographics, comorbidities, triage conditions, initial laboratory results, and primary ED diagnosis; critical ED interventions such as intubation, mechanical ventilation, and vasopressor use; and patient disposition including emergency operation, ICU admission, and admitting department (eg, surgical versus medical) were assessed from the EHR database. The comorbidities were assessed using Elixhauser's algorithm, revised by Quan et al in 2005 14 using International Classification of Disease (ICD‐10) code entry of the past outpatient and inpatient visits, as well as the index admission.
2.3. Statistical analyses
Categorical variables were reported using frequencies and proportions. Continuous variables were reported using median and interquartile range (IQR). t‐test, Wilcoxon's rank‐sum test, chi‐square test, and Fisher's exact test were performed as appropriate for comparisons between groups.
The association between prolonged ED stay and HAPU was modelled using multiple logistic regression. However, earlier discharge before reaching the 1‐month margin led to right censoring. Therefore, we also built Cox regression models to observe whether the results are consistent. Cox regression requires the proportionality assumption to be met. However, the effect size of prolonged ED stay should diminish within several days after admission, and thus, the assumption will not hold. Therefore, we applied extended Cox regression using time‐varying coefficients (TVCs) to model the changing effect size. Specifically, we estimated the hazard ratios (HRs) and their 95% confidence intervals (CIs) in three pre‐specified post‐admission intervals (first week, second week, and the rest of the month).15
For both the regression approaches, we used two different covariate sets for multivariable adjustment. The first one includes age, gender, type of visit, intubation, mechanical ventilation, vasopressor use, emergency operation, ICU admission, and admitting department (medical versus surgical). The second one is a full covariate set with additional covariates such as primary ED diagnosis; van Walraven comorbidity score16; modified early warning score (MEWS); initial consciousness; and laboratory results including white blood cell (WBC), haematocrit, prothrombin time in international normalised ratio (PT INR), platelet, blood urea nitrogen (BUN), creatinine, bilirubin, and albumin level. Because primary ED diagnosis has high cardinality, we used level coding in which labelling data are substituted by the average risk of outcome event.17
The goodness of fit of the logistic and the extended Cox regression models were tested using the Hosmer‐Lemeshow test and Gronnesby and Borgan test, respectively. The proportional hazard assumption was assessed by testing each covariate to observe whether they have a significant interaction with log‐transformed time.18 The results of the models were presented as odds ratios (ORs) with 95% CIs and HRs with 95% CIs. P‐values <0.05 were considered significant. All data handling and statistical analyses were performed using R‐packages version 3.5.1 (R Foundation for Statistical Computing, Vienna, Austria).
3. RESULTS
A total of 48 641 adult patients admitted through the ED were identified and included as the study population. Among the population, a total of 930 patients (1.9%) had HAPUs within a month. Table 1 shows the baseline characteristics of the patients grouped by ED LOS category. Most of the covariates except gender significantly differed by ED LOS. Longer ED LOS was associated with increased risk of HAPUs within a month of admission (P < .001 and P‐trend <.001).
Table 1.
Baseline characteristics of study population
| ≤6 h | 6‐12 h | 12‐24 h | >24 h | P | |
|---|---|---|---|---|---|
| (N = 15 640) | (N = 14 181) | (N = 10 304) | (N = 8516) | ||
| Age, categorised, N (%) | <.001 | ||||
| ≤45 y | 4251 (27.2%) | 2721 (19.2%) | 1802 (17.5%) | 1115 (13.1%) | |
| 46‐65 y | 4794 (30.7%) | 4332 (30.5%) | 3129 (30.4%) | 2451 (28.8%) | |
| 66‐75 y | 2989 (19.1%) | 2932 (20.7%) | 2225 (21.6%) | 1976 (23.2%) | |
| 76‐85 y | 2688 (17.2%) | 3080 (21.7%) | 2319 (22.5%) | 2177 (25.6%) | |
| >85 y | 918 (5.9%) | 1116 (7.9%) | 829 (8.0%) | 797 (9.4%) | |
| Age, median (IQR) | 61.0 (44.0‐75.0) | 66.0 (50.0‐77.0) | 67.0 (52.0‐78.0) | 69.0 (55.0‐79.0) | <.001 |
| Gender, N (%) | .052 | ||||
| Female | 7219 (46.2%) | 6551 (46.2%) | 4671 (45.3%) | 3795 (44.6%) | |
| Male | 8421 (53.8%) | 7630 (53.8%) | 5633 (54.7%) | 4721 (55.4%) | |
| Elixhauser comorbidities, N (%) | |||||
| Congestive heart failure | 826 (5.3%) | 928 (6.5%) | 579 (5.6%) | 564 (6.6%) | <.001 |
| Cardiac arrhythmia | 1405 (9.0%) | 1087 (7.7%) | 671 (6.5%) | 655 (7.7%) | <.001 |
| Valvular disease | 271 (1.7%) | 260 (1.8%) | 149 (1.4%) | 152 (1.8%) | .117 |
| Pulmonary circulation disorders | 173 (1.1%) | 231 (1.6%) | 193 (1.9%) | 229 (2.7%) | <.001 |
| Peripheral vascular disorders | 548 (3.5%) | 457 (3.2%) | 326 (3.2%) | 284 (3.3%) | .415 |
| Hypertension | 2660 (17.0%) | 2404 (17.0%) | 1643 (15.9%) | 1480 (17.4%) | .044 |
| Paralysis | 110 (0.7%) | 103 (0.7%) | 85 (0.8%) | 58 (0.7%) | .634 |
| Other neurological disorders | 793 (5.1%) | 914 (6.4%) | 750 (7.3%) | 659 (7.7%) | <.001 |
| Chronic pulmonary disease | 753 (4.8%) | 1089 (7.7%) | 831 (8.1%) | 849 (10.0%) | <.001 |
| Diabetes, uncomplicated | 1411 (9.0%) | 1433 (10.1%) | 1093 (10.6%) | 1121 (13.2%) | <.001 |
| Diabetes, complicated | 485 (3.1%) | 713 (5.0%) | 480 (4.7%) | 427 (5.0%) | <.001 |
| Hypothyroidism | 242 (1.5%) | 261 (1.8%) | 181 (1.8%) | 165 (1.9%) | .104 |
| Renal failure | 678 (4.3%) | 920 (6.5%) | 717 (7.0%) | 815 (9.6%) | <.001 |
| Liver disease | 570 (3.6%) | 960 (6.8%) | 876 (8.5%) | 806 (9.5%) | <.001 |
| Peptic ulcer disease excluding bleeding | 111 (0.7%) | 175 (1.2%) | 175 (1.7%) | 193 (2.3%) | <.001 |
| AIDS/HIV | 6 (0.0%) | 9 (0.1%) | 9 (0.1%) | 11 (0.1%) | .078 |
| Lymphoma | 221 (1.4%) | 287 (2.0%) | 241 (2.3%) | 293 (3.4%) | <.001 |
| Metastatic cancer | 497 (3.2%) | 814 (5.7%) | 760 (7.4%) | 996 (11.7%) | <.001 |
| Solid tumour without metastasis | 1867 (11.9%) | 2709 (19.1%) | 2193 (21.3%) | 2139 (25.1%) | <.001 |
| Rheumatoid arthritis or collagen vascular diseases | 178 (1.1%) | 295 (2.1%) | 194 (1.9%) | 186 (2.2%) | <.001 |
| Coagulopathy | 83 (0.5%) | 122 (0.9%) | 101 (1.0%) | 106 (1.2%) | <.001 |
| Obesity | 43 (0.3%) | 29 (0.2%) | 20 (0.2%) | 10 (0.1%) | .079 |
| Weight Loss | 37 (0.2%) | 53 (0.4%) | 25 (0.2%) | 28 (0.3%) | .106 |
| Fluid and electrolyte disorders | 227 (1.5%) | 323 (2.3%) | 268 (2.6%) | 283 (3.3%) | <.001 |
| Blood loss anaemia | 1 (0.0%) | 1 (0.0%) | 4 (0.0%) | 2 (0.0%) | .164 |
| Deficiency anaemia | 87 (0.6%) | 122 (0.9%) | 110 (1.1%) | 99 (1.2%) | <.001 |
| Alcohol abuse | 57 (0.4%) | 70 (0.5%) | 55 (0.5%) | 45 (0.5%) | .142 |
| Drug abuse | 91 (0.6%) | 51 (0.4%) | 53 (0.5%) | 51 (0.6%) | .026 |
| Psychoses | 117 (0.7%) | 56 (0.4%) | 52 (0.5%) | 27 (0.3%) | <.001 |
| Depression | 552 (3.5%) | 486 (3.4%) | 405 (3.9%) | 326 (3.8%) | .126 |
| Triage conditions, N (%) | |||||
| Injury‐related visit | 886 (5.7%) | 1159 (8.2%) | 834 (8.1%) | 445 (5.2%) | <.001 |
| Modified early warning score | 1.0 (1.0‐3.0) | 1.0 (1.0‐3.0) | 2.0 (1.0‐3.0) | 2.0 (1.0‐3.0) | <.001 |
| Altered mentality | 2358 (15.1%) | 1600 (11.3%) | 1118 (10.9%) | 959 (11.3%) | <.001 |
| Initial laboratory test results | |||||
| White blood cell count, 109/L, median (IQR) | 8.8 (6.6‐12.0) | 8.9 (6.5‐12.3) | 9.1 (6.7‐12.7) | 9.1 (6.4‐12.8) | <.001 |
| Haematocrit, %, median (IQR) | 39.7 (35.1‐43.7) | 38.4 (33.6‐42.6) | 37.6 (32.6‐42.0) | 35.8 (30.7‐40.5) | <.001 |
| Prothrombin time, INR, median (IQR) | 1.0 (1.0‐1.1) | 1.0 (1.0‐1.1) | 1.1 (1.0‐1.2) | 1.1 (1.0‐1.2) | <.001 |
| Platelet, 109/L, median (IQR) | 218.0 (174.0‐265.0) | 215.0 (165.0‐271.0) | 212.0 (158.0‐269.0) | 207.0 (148.0‐273.0) | <.001 |
| Blood urea nitrogen, mg/dL, median (IQR) | 15.0 (11.0‐20.0) | 16.0 (12.0‐23.0) | 17.0 (12.0‐25.0) | 18.0 (12.0‐29.0) | <.001 |
| Creatinine, mg/dL, median (IQR) | 0.8 (0.6‐1.0) | 0.8 (0.6‐1.1) | 0.8 (0.6‐1.1) | 0.8 (0.6‐1.3) | <.001 |
| Bilirubin, total, mg/dL, median (IQR) | 0.6 (0.4‐0.9) | 0.7 (0.4‐1.0) | 0.7 (0.4‐1.1) | 0.7 (0.5‐1.1) | <.001 |
| Albumin, g/dL, median (IQR) | 4.0 (3.5‐4.3) | 3.9 (3.3‐4.3) | 3.8 (3.3‐4.2) | 3.6 (3.1‐4.1) | <.001 |
| ED interventions, N (%) | |||||
| Endotracheal intubation | 913 (5.8%) | 391 (2.8%) | 196 (1.9%) | 108 (1.3%) | <.001 |
| Mechanical ventilation | 753 (4.8%) | 388 (2.7%) | 199 (1.9%) | 101 (1.2%) | <.001 |
| Vasopressors | 2 (0.0%) | 76 (0.5%) | 582 (5.6%) | 654 (7.7%) | <.001 |
| Adjustment scores | |||||
| Admission diagnosis (level coded) | −0.2 (−1.4‐0.2) | −0.2 (−1.0‐0.2) | −0.2 (−0.9‐0.3) | −0.1 (−0.6‐0.3) | <.001 |
| Van Walraven score | 0.0 (0.0‐5.0) | 2.0 (0.0‐7.0) | 4.0 (0.0‐9.0) | 4.0 (0.0‐11.0) | <.001 |
| Disposition after ED, N (%) | |||||
| ICU admission | 3747 (24.0%) | 1670 (11.8%) | 692 (6.7%) | 260 (3.1%) | <.001 |
| Operating room (emergency operation) | 718 (4.6%) | 508 (3.6%) | 216 (2.1%) | 85 (1.0%) | <.001 |
| Admitting department | <.001 | ||||
| Medical | 10 116 (64.7%) | 9694 (68.4%) | 7825 (75.9%) | 7227 (84.9%) | |
| Surgical | 5524 (35.3%) | 4487 (31.6%) | 2479 (24.1%) | 1289 (15.1%) | |
| Length of stay excluding ED, days, median (IQR) | 6.0 (4.0‐12.0) | 7.0 (4.0‐12.0) | 7.0 (5.0‐13.0) | 8.0 (5.0‐15.0) | <.001 |
| HAPU within a month after admission | 275 (1.8%) | 247 (1.7%) | 194 (1.9%) | 214 (2.5%) | <.001/P‐trend<.001 |
Table 2 describes the characteristics of the HAPU lesions at initial recognition. The median size was 10.0 cm2 (IQR, 2.0‐35.0 cm2), and the most common anatomic location and stage was the coccygeal area (50.1%) and stage 2 (54.6%), respectively. The most common presumed cause was altered mental status (234, 25.2%). Most of the initial characteristics did not differ by ED LOS except prolonged operation (P = .047) or haemodynamic instability (P = .019).
Table 2.
Characteristics of the hospital‐acquired pressure ulcer (HAPU) lesions at initial encounters
| Overall (N = 930) | Grouped by ED LOS | P | ||||
|---|---|---|---|---|---|---|
| ≤6 h | 7‐12 h | 13‐24 h | >24 h | |||
| (N = 275) | (N = 247) | (N = 194) | (N = 214) | |||
| Location of PUa | .431 | |||||
| Coccyx | 466 (50.1%) | 133 (48.4%) | 126 (51.0%) | 103 (53.1%) | 104 (48.6%) | |
| Hip | 162 (17.4%) | 49 (17.8%) | 38 (15.4%) | 28 (14.4%) | 47 (22.0%) | |
| Heel | 42 (4.5%) | 7 (2.5%) | 15 (6.1%) | 11 (5.7%) | 9 (4.2%) | |
| Trochanter | 38 (4.1%) | 10 (3.6%) | 12 (4.9%) | 9 (4.6%) | 7 (3.3%) | |
| Vertebral | 33 (3.5%) | 8 (2.9%) | 9 (3.6%) | 5 (2.6%) | 11 (5.1%) | |
| Ear | 24 (2.6%) | 9 (3.3%) | 8 (3.2%) | 3 (1.5%) | 4 (1.9%) | |
| Outer ankle | 18 (1.9%) | 6 (2.2%) | 5 (2.0%) | 3 (1.5%) | 4 (1.9%) | |
| Sacrum | 18 (1.9%) | 6 (2.2%) | 2 (0.8%) | 5 (2.6%) | 5 (2.3%) | |
| Nose | 16 (1.7%) | 10 (3.6%) | 4 (1.6%) | 2 (1.0%) | 0 (0.0%) | |
| Scapula | 15 (1.6%) | 5 (1.8%) | 4 (1.6%) | 5 (2.6%) | 1 (0.5%) | |
| Others | 98 (10.5%) | 32 (11.6%) | 24 (9.7%) | 20 (10.3%) | 22 (10.3%) | |
| Size of PU | ||||||
| Width | 3.0 (1.5‐6.0) | 4.0 (1.0‐6.0) | 3.0 (2.0‐6.0) | 3.0 (1.5‐8.0) | 3.0 (1.0‐5.0) | .119 |
| Height | 3.0 (1.5‐5.0) | 3.0 (1.0‐5.0) | 3.0 (2.0‐5.0) | 3.0 (1.0‐6.0) | 3.0 (1.5‐5.0) | .582 |
| Area (width * height) | 10.0 (2.0‐35.0) | 11.0 (2.0‐36.0) | 11.0 (3.0‐27.5) | 10.0 (2.0‐48.0) | 9.0 (1.5‐25.0) | .400 |
| Initial stage at first encounter | .732 | |||||
| Stage 1 | 415 (44.7%) | 130 (47.3%) | 115 (46.7%) | 85 (43.8%) | 85 (39.7%) | |
| Stage 2 | 507 (54.6%) | 143 (52.0%) | 129 (52.4%) | 108 (55.7%) | 127 (59.3%) | |
| Stage 3 | 7 (0.8%) | 2 (0.7%) | 2 (0.8%) | 1 (0.5%) | 2 (0.9%) | |
| Presumed causes | ||||||
| Altered mental status | 234 (25.2%) | 73 (26.5%) | 68 (27.5%) | 41 (21.1%) | 52 (24.3%) | .428 |
| Prolonged surgery | 193 (20.8%) | 66 (24.0%) | 50 (20.2%) | 46 (23.7%) | 31 (14.5%) | .047 |
| Oedema | 181 (19.5%) | 52 (18.9%) | 53 (21.5%) | 42 (21.6%) | 34 (15.9%) | .389 |
| Incontinence | 176 (18.9%) | 46 (16.7%) | 49 (19.8%) | 36 (18.6%) | 45 (21.0%) | .652 |
| Motor/sensory dysfunction | 164 (17.6%) | 52 (18.9%) | 45 (18.2%) | 34 (17.5%) | 33 (15.4%) | .780 |
| Haemodynamic instability | 147 (15.8%) | 49 (17.8%) | 48 (19.4%) | 30 (15.5%) | 20 (9.3%) | .019 |
| Device‐related | 73 (7.8%) | 29 (10.5%) | 20 (8.1%) | 10 (5.2%) | 14 (6.5%) | .155 |
| Others | 377 (40.5%) | 100 (36.4%) | 98 (39.7%) | 78 (40.2%) | 101 (47.2%) | .111 |
Ten most common anatomical locations.
Univariable logistic and Cox regression were performed. The crude OR and HR for HAPUs increased to 1.44 (95% CI, 1.20‐1.72) and 1.21 (95% CI, 1.02‐1.45), respectively, in ED LOS ≥ 24 hours relative to ED LOS < 6 hours (Table 3).
Table 3.
Crude relationship between ED LOS and hospital‐acquired pressure ulcer (HAPU) without any adjustment
| Logistic regression | Cox regression | |||
|---|---|---|---|---|
| OR | P | HR | P | |
| ≤6 h (Ref.) | 1.00 (1.00‐1.00) | N/A | 1.00 (1.00‐1.00) | N/A |
| 6‐12 h | 0.99 (0.83‐1.18) | .913 | 1.00 (0.84‐1.18) | .974 |
| 12‐24 h | 1.07 (0.89‐1.29) | .462 | 1.00 (0.84‐1.21) | .967 |
| >24 h | 1.44 (1.20‐1.72) | <.001 | 1.21 (1.02‐1.45) | .033 |
Multivariable adjustment was performed with multiple logistic and extended Cox regression models (Figure 1, Table 4). The complete description of the parameters is given in the Supplementary Tables 1 to 4. In the logistic regression models, ED LOS ≥12 and ≥24 hours were associated with increased risk of HAPUs with ORs of 1.50 (95% CI, 1.23‐1.84) and 2.27 (95% CI, 1.85‐2.79) relative to ED LOS <6 hours, respectively, when adjusted with covariate set 1 and 1.30 (95% CI, 1.05‐1.60) and 1.80 (95% CI, 1.45‐2.23) relative to ED LOS < 6 hours, respectively, when adjusted with covariate set 2.
Figure 1.
The association between prolonged emergency department (ED) stay and the risk of new pressure ulcer assessed using logistic and extended Cox regression with time‐varying coefficients. A, Logistic regression model, B, piecewise Cox regression model
Table 4.
Adjusted effect sizes of prolonged ED LOS on the risk of developing hospital‐acquired pressure ulcer (HAPU)
| Logistic regression | Extended Cox regression with TVC | |||||||
|---|---|---|---|---|---|---|---|---|
| First week | Second week | Rest of the month | ||||||
| OR | P | HR | P | HR | P | HR | p | |
| Covariate set #1 | ||||||||
| ≤6 h (Ref.) | 1.00 (1.00‐1.00) | N/A | 1.00 (1.00‐1.00) | N/A | 1.00 (1.00‐1.00) | N/A | 1.00 (1.00‐1.00) | N/A |
| 6‐12 h | 1.20 (1.00‐1.44) | .044 | 1.21 (0.95‐1.56) | .126 | 0.99 (0.69‐1.41) | .947 | 1.22 (0.86‐1.74) | .266 |
| 12‐24 h | 1.50 (1.23‐1.84) | <.001 | 1.51 (1.15‐1.99) | .003 | 0.94 (0.63‐1.42) | .780 | 1.42 (0.97‐2.08) | .069 |
| >24 h | 2.27 (1.85‐2.79) | <.001 | 2.24 (1.69‐2.95) | <.001 | 1.50 (1.02‐2.21) | .039 | 1.55 (1.04‐2.31) | .031 |
| Covariate set #2 | ||||||||
| ≤6 h (Ref.) | 1.00 (1.00‐1.00) | N/A | 1.00 (1.00‐1.00) | N/A | 1.00 (1.00‐1.00) | N/A | 1.00 (1.00‐1.00) | N/A |
| 6‐12 h | 1.09 (0.91‐1.32) | .351 | 1.17 (0.91‐1.51) | .214 | 0.91 (0.63‐1.30) | .594 | 1.07 (0.75‐1.52) | .723 |
| 12‐24 h | 1.30 (1.05‐1.60) | .015 | 1.42 (1.07‐1.88) | .015 | 0.85 (0.56‐1.29) | .454 | 1.24 (0.84‐1.83) | .274 |
| >24 h | 1.80 (1.45‐2.23) | <.001 | 1.92 (1.44‐2.57) | <.001 | 1.31 (0.88‐1.94) | .186 | 1.27 (0.84‐1.90) | .255 |
The extended Cox regression models showed that the risk is transient up to a week with increased HR of 1.51 (95% CI, 1.15‐1.99) and 2.24 (95% CI, 1.69‐2.95) relative to ED LOS < 6 hours, respectively, using covariate set 1 and 1.42 (95% CI, 1.07‐1.88) and 1.92 (95% CI, 1.44‐2.57) relative to ED LOS < 6 hours, respectively, using covariate set 2.
4. DISCUSSION
In this study, we assessed the association between prolonged ED LOS and HAPU using a prospective surveillance dataset. We found that prolonged ED LOS over 12 hours is an independent risk factor for HAPU within a week of admission. To the best of our knowledge, this is the first report elucidating the relationship between the two entities. The results of the study suggest that reducing ED LOS should be a goal in a multifaceted approach to prevent HAPUs in hospitals.
Previous studies could not find any meaningful relationship between ED LOS and HAPUs. Baumgarten et al. assessed the impact of various exposures, including longer ED LOS, night or weekend admission, potentially immobilising procedures, and admission to an ICU in a nested case‐control study of patients aged ≥65 years admitted through the ED (Case, HAPU, N = 195; Control, No HAPU, N = 597).9 In the study, mean ED LOS was not significantly different between mean 7.2 hours (standard deviation [SD], 3.1) in cases and 7.4 hours (SD, 3.4) in controls (P = .607).10 Scott et al. retrospectively analysed 314 surgical patients to identify risk factors associated with HAPU. In the analysis, 43 patients had HAPU, and the length of time spent on a trolley in the accident department and ED was not significantly different in the group. Obviously, the primary objective of the studies was screening potential risk factors rather than proving or disproving the hypothesis that ED LOS is associated with HAPU.
Incidence of HAPU can be reduced by implementing an intensive tracking system.19 Incidence tracking system through EMR may facilitate early identification, initiation of effective care, and staff education.19 Structured risk assessment for HAPUs, including skin and tissue status, may prevent HAPUs.20 However, in the crowded ED setting, it is not easy to perform a complete risk assessment. Skin care or repositioning can be applied in the ED setting. Avoiding the positioning of patients on an area of erythema20 and placing thin foam or breathable dressings under a medical device may prevent HAPU incidence.21
This study has several limitations. First, we could not adjust for several known risk factors of PUs, including incontinence, smoking, and pressure‐reducing bed surface, because such information was not available in patients without PU event.22 Second, we used ICD codes to assess underlying comorbidities, which can lead to misclassification of issues. Third, information about the events following ED stay was not available. Such information can be useful to gain a comprehensive view of the risks of HAPU after the exposure. Finally, this is a single‐centre study that can be a concern for generalisability. For example, in non‐overcrowded EDs with adequate resources, the impact of prolonged ED stay could be smaller.
5. CONCLUSIONS AND IMPLICATIONS
Prolonged ED stay over 12 hours is an independent risk factor for HAPUs within a week after admission. Reducing ED LOS should be pursued as a goal in a multifaceted approach to reduce PUs in hospitalised patients.
CONFLICT OF INTEREST
The authors declare no potential conflict of interest.
Supporting information
Table S1–S4.
ACKNOWLEDGEMENTS
This work was supported by grant from the Seoul National University Bundang Hospital Research Fund, No. 13‐2017‐015.
Han D, Kang B, Kim J, et al. Prolonged stay in the emergency department is an independent risk factor for hospital‐acquired pressure ulcer. Int Wound J. 2020;17:259–267. 10.1111/iwj.13266
Funding information Seoul National University Bundang Hospital Research Fund, Grant/Award Number: 13‐2017‐015
REFERENCES
- 1. Edsberg LE, Black JM, Goldberg M, McNichol L, Moore L, Sieggreen M. Revised national pressure ulcer advisory panel pressure injury staging system. J Wound Ostomy Cont Nurs. 2016;43(6):585‐597. 10.1097/won.0000000000000281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Mervis JS, Phillips TJ. Pressure ulcers: pathophysiology, epidemiology, risk factors, and presentation. J Am Acad Dermatol. 2019;81:881‐890. 10.1016/j.jaad.2018.12.069. [DOI] [PubMed] [Google Scholar]
- 3. Kaltenthaler E, Withfield WSJ, Akehurst RL, Paisley S. UK, USA and Canada: how do their pressure ulcer prevalence and incidence data compare? J Wound Care. 2001;10(1):530‐535. 10.12968/jowc.2001.10.1.26039. [DOI] [PubMed] [Google Scholar]
- 4. Lyder CH, Wang Y, Metersky M, et al. Hospital‐acquired pressure ulcers: results from the national medicare patient safety monitoring system study. J Am Geriatr Soc. 2012;60(9):1603‐1608. 10.1111/j.1532-5415.2012.04106.x. [DOI] [PubMed] [Google Scholar]
- 5. Pines JM, Hilton JA, Weber EJ, et al. International perspectives on emergency department crowding. Acad Emerg Med. 2011;18(12):1358‐1370. 10.1111/j.1553-2712.2011.01235.x. [DOI] [PubMed] [Google Scholar]
- 6. Somma S, Paladino L, Vaughan L, Lalle I, Magrini L, Magnanti M. Overcrowding in emergency department: an international issue. Intern Emerg Med. 2015;10(2):171‐175. 10.1007/s11739-014-1154-8. [DOI] [PubMed] [Google Scholar]
- 7. Brouns SH, Stassen PM, Lambooij SL, Dieleman J, Vanderfeesten IT, Haak HR. Organisational factors induce prolonged emergency department length of stay in elderly patients—A retrospective cohort study. PLoS One. 2015;10(8):e0135066. 10.1371/journal.pone.0135066. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Jeanmonod D, Jeanmonod R. Vignettes in Patient Safety. London: IntechOpen; 2018;2:57‐72. 10.5772/intechopen.69243 [DOI] [Google Scholar]
- 9. Baumgarten M, Margolis DJ, Localio AR, et al. Extrinsic risk factors for pressure ulcers early in the hospital stay: a nested case‐control study. J Gerontol A Biol Sci Med Sci. 2008;63(4):408‐413. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Scott EM. Hospital‐acquired pressure sores in surgical patients. J Wound Care. 1998;7(2):76‐79. 10.12968/jowc.1998.7.2.76. [DOI] [PubMed] [Google Scholar]
- 11. Goldstone LA, Goldstone J. The Norton score: an early warning of pressure sores? J Adv Nurs. 1982;7(5):419‐426. 10.1111/j.1365-2648.1982.tb00258.x. [DOI] [PubMed] [Google Scholar]
- 12. Baumgarten M, Margolis DJ, Orwig DL, et al. Pressure ulcers in elderly patients with hip fracture across the continuum of care. J Am Geriatr Soc. 2009;57(5):863‐870. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Liu P, Shen W‐Q, Chen H‐L. The incidence of pressure ulcers in the emergency department: a metaanalysis. Wounds Compend Clin Res Pract. 2016;29:14‐19. [PubMed] [Google Scholar]
- 14. Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD‐9‐CM and ICD‐10 administrative data. Med Care. 2005;43(11):1130‐1139. 10.1097/01.mlr.0000182534.19832.83. [DOI] [PubMed] [Google Scholar]
- 15. Therneau T, Crowson C, Vignettes AE. Using time dependent covariates and time dependent coefficients in the Cox model. R package Survival Vignettes, The Comprehensive R Archive Network. 2019. Available at: http://cran.r-project.org/web/packages/survival/vignettes/timedep.pdf. Accessed November 5, 2019.
- 16. van Walraven C, Austin PC, Jennings A, Quan H, Forster AJ. A modification of the Elixhauser comorbidity measures into a point system for hospital death using administrative data. Med Care. 2009;47(6):626‐633. 10.1097/mlr.0b013e31819432e5. [DOI] [PubMed] [Google Scholar]
- 17. Moeyersoms J, Martens D. Including high‐cardinality attributes in predictive models: a case study in churn prediction in the energy sector. Decis Support Syst. 2015;72:72‐81. 10.1016/j.dss.2015.02.007. [DOI] [Google Scholar]
- 18. Grambsch P, Therneau TM. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika. 1994;81:515‐526. 10.1093/biomet/81.3.515. [DOI] [Google Scholar]
- 19. Zaratkiewicz S, Whitney JD, Lowe JR, Taylor S, O'Donnell F, MintonFoltz P. Development and implementation of a hospital‐acquired pressure ulcer incidence tracking system and algorithm. J Healthc Qual. 2010;32(6):44‐51. 10.1111/j.1945-1474.2010.00076.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. National Pressure Ulcer Advisory Panel . Pressure injury prevention points. 2016. https://cdn.ymaws.com/npuap.org/resource/resmgr/1a._pressure‐injury‐preventi.pdf. Accessed October 20, 2019.
- 21. Kuo CY, Wootten CT, Tylor DA, Werkhaven JA, Huffman KF, Goudy SL. Prevention of pressure ulcers after pediatric tracheotomy using a Mepilex Ag dressing. Laryngoscope. 2013;123(12):3201‐3205. 10.1002/lary.24094. [DOI] [PubMed] [Google Scholar]
- 22. Çakmak S, Gül Ü, Özer S, Yiit Z, Gönü M. Risk factors for pressure ulcers. Adv Skin Wound Care. 2009;22(9):412‐415. 10.1097/01.asw.0000360256.99980.84. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1–S4.